Tesla scored a significant victory in China by invalidating a 20-year-old AI patent (CN100399329C) that smartphone-maker Xiaomi had licensed for years. The patent, titled “Mobile Intelligent Navigation Engine System,” utilizes natural language processing (NLP) for mobile services, and was initially filed before smartphones existed. Tesla’s challenge relied on earlier disclosures, including a 2004 U.S. patent (US20040167875A1), which taught similar NLP-based navigation. Citing obviousness, China’s Patent Office struck down the patent. This decision has sparked debate in China over increasingly strict inventiveness standards.
Tesla’s challenge relied on earlier disclosures, including a 2004 U.S. patent (US20040167875A1) that taught similar NLP-based navigation. The China National Intellectual Property Administration (CNIPA) invalidated the patent on the grounds of obviousness, citing prior art that rendered the patented invention non-inventive under Chinese law.
This decision has sparked debate in China over increasingly strict standards for inventiveness, particularly for software and AI-related patents. The Tesla ruling is seen as part of a broader trend by CNIPA to establish higher thresholds of non-obviousness, aligning more closely with international norms and curbing overly broad filings.
Xiaomi’s Licensing Misstep
In 2019, a Chinese SME secured a licensing deal with Xiaomi – “the only publicly disclosed instance of Xiaomi paying royalties to a domestic SME”. Xiaomi even launched straw-man invalidation challenges during negotiations, but ultimately “conceded the patent’s validity and signed the agreement”.
In short, Xiaomi paid royalties for roughly five years despite the patent’s obviousness vulnerability.
This case serves as a reminder that licensing without a rigorous prior art search can become costly, both in terms of reputation and finances.
The Power of a Thorough Prior Art Search
Tesla’s win vividly illustrates what a strong search can achieve. To avoid surprises like this, companies must invest in comprehensive search strategies, including:
- Multi-Database (MDS) Coverage: Search across multiple databases. Use comprehensive global collections of patents and technical literature. In practice, this means searching various databases and even non-patent sources (journal articles, standards, earlier products). Incomplete or outdated data can hide key art, so the breadth of coverage is critical.
- Multilingual Expertise: Don’t rely solely on English or machine translation. Skilled IP searchers fluent in the source languages can spot nuances in foreign disclosures. Human review of Chinese, English, or other publications helps uncover prior art that simple keyword queries might miss.
- Smart Search Strategies: A one-shot keyword search is seldom enough. Searcher should combine approaches – using Boolean queries, patent classification codes (IPC/CPC), citation trees, and semantic or concept-based searches – to connect the dots across technology domains.
- Iterative and Creative Process: Quality searches take time and creativity. Instead of a quick, literal search, analysts should iteratively refine queries, explore synonyms, and map related fields. Rushing can miss critical hits – The goal is to reveal hidden connections and inventive-step risks rather than merely matching surface terms.
In short, a strong search casts a wide net, utilizes human expertise, and systematically explores all relevant information. Each of the above elements, if applied during Xiaomi’s licensing, might have unearthed the fatal prior art and altered the outcome.
Our Prior Art Search Expertise
At Evalueserve IP and R&D, we apply all these best practices to safeguard clients’ IP. Our approach includes:
- Access to Global Multiple Patent Databases as well as non-patent repositories
We combine databases across jurisdictions (China, US, Europe, WIPO) with extensive non-patent collections to surface buried or overlooked art.
- Multilingual Technical Analysts
Our search team includes IP experts fluent in Chinese, German, Japanese, and other languages. They read and interpret source documents directly, ensuring subtle disclosures aren’t lost in translation.
- Rigorous Search Methodology
We employ a layered process that combines keyword searches, classification-code queries, citation analysis, semantic mapping, and context-driven refinement.
- Creative Problem-Solvers.
Our searchers are trained to think broadly, following citation trails and related technologies rather than just literal terms. This creative, contextual approach often uncovers “needle in a haystack” references that generic searches miss.
- Augmented by AI-Powered Tools for Deeper Insights
In line with our Mind+Machine philosophy, we integrate AI-driven tools to complement expert analysis and insights. These tools enhance concept recognition, semantic clustering, and pattern identification across massive datasets, accelerating detection of prior art that might elude traditional methods.
By integrating top-tier tools with experienced human analysis, we maximize the likelihood of identifying any potentially threatening prior art. We help clients do just that – strengthening their IP with defensible, evidence-backed insights.
Why Thorough Prior Art Search Matters
A proactive, exhaustive search is more than a box-check; it’s business-critical. Effective prior art investigations:
- Ensure Patent Quality
They verify that an invention truly meets the standards of novelty and non-obviousness. In fact, thorough searches prevent the patenting of existing ideas, thereby upholding the integrity of patent systems. Patents backed by solid art searches are far less likely to be invalidated later, preserving their value and enforceability.
- Reduce Litigation Risk
By uncovering weaknesses upfront, companies avoid nasty surprises. Exhaustive searches protect inventors from future legal complications and support more vigorous enforcement if needed.
- Support Licensing and Investment
Well-researched patents are more attractive to potential buyers in mergers and acquisitions. Detailed search reports demonstrate patent novelty and limitability, reducing perceived risk for licensees and investors. A strong patent portfolio can command better licensing terms and valuation, directly affecting a company’s bottom line.
In Xiaomi’s case, a thorough search would have made Xiaomi’s deal fairer or prevented it altogether. In your case, the right search strategy means confident licensing, no invalidation threats, and a more explicit focus on genuine innovation.
Proactive IP Strategy – Let’s Partner Up
Don’t wait for an adversary’s challenge to uncover hidden art. Proactively searching for prior art is the best defence against costly litigation and bad licensing deals.
Whether you’re preparing for filing, licensing, M&A, or litigation, our search experts can provide the clarity and confidence you need to move forward.
Let’s start with a quick consultation to review your most valuable assets or upcoming transactions. Together, we’ll build a defensive, forward-looking IP strategy – so you can invest in innovation with confidence.
You could explore our search service here: Home | Searchstream
